/Phase2.2MISCRPackage

Primary LanguageRApache License 2.0Apache-2.0

Investigating the epidemiology of MIS-C

Here we provide the steps for running Phase2.2 MISC R package. This package specifically requres Phase2.2 data.

Install and run

1. Install the R pacakge

In R, install and load the latest version of Phase2.2MISCRPackage from GitHub with remotes and run:

remotes::install_github('covidclinical/Phase2.2MISCRPackage',
                        subdir = 'FourCePhase2.2MISC',
                        upgrade = FALSE)

library(FourCePhase2.2MISC)

2. Run the analysis.

To run the analysis you will have to run the function 'FourCePhase2.2MISC::runAnalysis'. Before running it update the arguments accordingly to your site.

  • dir.input will be the path to the folder where you have the 2.2 MISC files. Note that the name of files should be the same than the regular 2.2 files
  • dir.output will be the folder where the results will be saved.
  • obfuscation: if you do not need to apply obfuscation to your site, leave it as FALSE, if not change it to the numeric value (e.g., 3)
  • raceAvailable: set it to FALSE if you do not have these information available.
  • country: set it to your country (e.g., "us", "france", "spain", "uk")
  • data_updated_date: change it to the date when your data was generated.
  • dateFormat: this is set to the dateFormat used in 4CE, change it if you have your dates in another format.
  • verbose: leave it as TRUE to get the log file
FourCePhase2.2MISC::runAnalysis( dir.input = "/path_input_2.2MISCdata/",
                                 dir.output = "/path_to_folder_to_save_output/",
                                 obfuscation = FALSE,
                                 raceAvailable= TRUE,
                                 country = "yourCountry",
                                 data_update_date = "2022-06-01",
                                 dateFormat = "%Y-%m-%d",
                                 verbose = TRUE)

Note that if you run the function multiple times you will get a warning saying that the output directory already exists. You can ignore the warning since the output files generated are overwritten each time.

Outputs

What outputs are generated?

  • figures folder: contains visual summaries of the cohort counts

    • Age distribution
    • Hospitalization length distribution
    • Number of patients in ICU per variant
    • Patient counts per month

  • QC folder:

    • SITEID_ICDdiagnosisCodes.RData: contains the ICD codes available in the patient population
    • MISC_logs_QC.txt: contains the runtime logs and warnings from running the package

  • SITEID_table1.txt: human readable Table 1 (demographics & comorbidities)

  • SITEID_table2.txt: human readable Table 2 (lab values)

  • SITEID_table3.txt: human readable Table 3 (diagnoses)

  • SITEID_table1Categorical.RData: raw, obfuscated, aggregate counts for Table 1 categorical variables (demographics & comorbidities)

  • SITEID_table1Continuous.RData: raw, obfuscated, aggregate counts for Table 1 continuous variables (demographics & comorbidities)

  • SITEID_table2AtAdmission.RData: raw, obfuscated, aggregate counts for Table 2 laboratory values at admission

  • SITEID_table2DuringAdmission.RData: raw, obfuscated, aggregate counts for Table 2 laboratory values during admission

  • SITEID_table3.RData: raw, obfuscated, aggregate counts for Table 3 (diagnoses)

3. Submit

Please share the generated output folder with us (@Simran Makwana, @Alba Gutierrez ) via the #pediatrics Slack channel. If you run into any problem adapting this code to your data, let us know via Slack.

Thank you very much for your contribution!